1.1. Field of the Invention
The present invention relates generally to the field of molecular biology. More particularly, certain embodiments concern methods and compositions comprising DNA segments and proteins derived from bacterial species. More particularly, the invention provides fnbA nucleic acid and FnBPA amino acid compositions from Staphylococcus aureus. Also disclosed are peptide epitopes and protein sequences comprising site-specifically-modified or truncated fibronectin (Fn) binding site domains, and antibodies derived from immunization of animals with these peptide epitopes and binding site domains. Various methods for making and using these antibodies, peptides and DNA segments, peptides and nucleic acid segments encoding modified ligand binding site domains, and native and synthetic proteins are disclosed, such as, for example, the use of antibodies and/or DNA segments as diagnostic probes and templates for protein production, and the use of antibodies, proteins, fusion protein carriers, peptides and nucleic acid segments in various pharmacological and immunological applications.
1.2. Description of the Related Art
1.2.1 MSCRAMMs
Bacterial adherence to host tissue involves specific microbial surface adhesins of which a subfamily termed MSCRAMMs specifically recognize extracellular matrix (ECM) components. Many pathogenic bacteria have been shown to specifically recognize and bind to various components of the extracellular matrix in an interaction which appears to represent a host tissue colonization mechanism. This adherence involves a group of bacterial proteins termed MSCRAMMs (microbial surface components recognizing adhesive matrix molecules) (Patti et al., 1994; Patti and Höök, 1994).
Several Fn binding MSCRAMMs have been isolated and characterized from different Gram-positive bacteria. Genes encoding Fn binding MSCRAMMs from Staphylococcus aureus (Signäs et al., 1989), Streptococcus pyogenes (Talay et al., 1994; Hansky et al., 1992) and Streptococcus dysgalactiae (Lindgren et al., 1993) have been cloned and sequenced. The deduced amino acid sequences revealed 60–100 kDa proteins with very similar structural organization. The N-terminal signal sequence is followed by a long stretch of unique sequence which in some cases is interrupted by two copies of an approximately 30 amino acid long segment. The ligand binding site is located just N-terminal of a proline-rich domain, which is believed to anchor the proteins in the cell wall. This domain is followed by the sequence LPXTGX (SEQ ID NO: 1) which is a cell wall targeting signal (Schneewind et al., 1995), a stretch of hydrophobic residues representing a trans-membrane unit and a short C-terminal cytoplasmic domain containing a cluster of positively charged residues. The primary Fn-binding sites on these MSCRAMMs consist of 30–42 amino acid long motifs repeated 3–4 times, and most of the repeated units contain a consensus sequence (Lindgren et al., 1993; McGavin et al., 1993). This domain is composed of a unit of 37–40 amino acids, repeated three or four times (FIG. 1).
Recombinant proteins corresponding to the repeat regions from the different Fn binding MSCRAMMs are all capable of inhibiting the binding of Fn to different Gram-positive bacteria, including S. aureus, S. dysgalactiae and S. pyogenes (Joh et al., 1994). Furthermore, studies using individual synthetic peptides revealed that a number of the repeat units retain Fn-binding activity, and interfere with binding of Fn by all of the Gram-positive species tested. These data suggest that the binding sites in Fn for the different MSCRAMMs are either overlapping or closely spaced on the matrix protein.
The repeat regions have been overexpressed as recombinant fusion proteins in Escherichia coli where the recombinant Fn binding domains (rFnBD) are linked to a stretch of histidine residues which are utilized for affinity purification of the rFnBD proteins. These proteins have been designated as rFnBD-D, rFnBD-A, rFnBD-B, and rFnBD-F, respectively FIG. 1. The rFnBDs were found to exhibit similar binding kinetics and dissociation constants; for example, the dissociation constants of the four recombinant proteins binding to porcine Fn was determined by biosensor analysis to be in the low nM range with the dominant dissociation rates varying between 1×10−1 and 6×10−1·s−1. Additionally, the recombinant proteins have been shown to have cross-species specificity and inhibit binding of Fn to many different bacterial cells (Joh et al., 1994).
The repeated units of the Fn binding domains of the different MSCRAMMs are strikingly similar, and appear to contain a consensus sequence (McGavin et al., 1991; House-Pompeo et al., 1996). The repeat units have a high number of acidic residues, and there are conserved hydrophobic and acidic residues at certain positions. Overall there is a high degree of sequence similarity between repeated units in a specific MSCRAMM as well as between MSCRAMMs from different species. Synthetic peptides, analogous to the repeated units, also bind Fn, and by amino acid substitution in these peptides it has been determined that all conserved residues are not needed for Fn binding (McGavin et al., 1991).
Fn is a disulfide-linked dimeric glycoprotein that is found in a soluble form in body fluids and a fibrillar form in the extracellular matrix. The primary biological function of Fn appears to be related to its ability to serve as a substrate for the adhesion of animal cells. This adhesion is mediated by a family of dimeric receptors which recognize and bind to specific sites in the central part of Fn. The primary binding sites in Fn for MSCRAMMs from Gram-positive bacteria has been localized to the Fn NH2-terminal domain (N29) (Mosher and Proctor, 1980; Speziale et al., 1984). This domain is composed of five type I modules which are about 45 amino acids in length. The structure of N29 is a series of anti-parallel β-sheets stabilized by several disulfide bonds interspersed at regular intervals in the sequences (Potts and Campbell, 1994; Venyaminov et al., 1983). The ability to bind Fn is located exclusively within the C-terminal 20 amino acids of each D-motif (Huff et al., 1994; McGavin et al., 1993; McGavin et al., 1991). These amino acids contain the sequence GG(X3,4)(I/V)DF, which is present in repeated motifs of other Fn-binding adhesins, and within the Fn-binding A2 motif of S. dysgalactiae FnBA, changes to either of the GG or IDF sequences resulted in loss of Fn-binding (McGavin et al. 1993).
The S. aureus Fn-binding MSCRAMM contains an additional ligand binding site in an approximately 30 amino acid long segment which has been designated Du that encompasses the consensus sequence and is located N-terminal of the repeat region. This segment can also interact with Fn and its N-terminal domain (designated N29) (Jönsson, 1992).
S. aureus possesses two tandem fnb genes, encoding Fn-binding proteins FnBPA and FnBPB (Jönsson et al., 1991; Signäs et al., 1989), each of which possesses three consecutive 37- or 38 amino acid D-motifs, designated D1, D2, and D3. In tandem, these motifs comprise a high affinity Fn-binding domain, D1–3. Synthetic peptides representing each motif are also individually capable of low affinity Fn-binding, and can competitively inhibit Fn-binding to S. aureus (Huff et al., 1994; Signäs et al., 1989).
1.2.2 Attempts to Generate Antibodies that Block Fn Binding have Failed
In all of the Fn-binding MSCRAMMs identified so far, the primary ligand binding sites have been located to domains composed of a 37–42 amino acid motif repeated 3–5 times (McGavin et al., 1993). Unfortunately, attempts to generate blocking antibodies employing both synthetic peptides and different forms of the D1–3 immunogen have been largely unsuccessful (Ciborowski et al., 1992; Rozalska et al., 1994; Speziale et al., 1996). Previous attempts to generate high affinity antibodies that could block S. aureus binding to Fn have had little success. For example, when rabbit polyclonal antibodies were generated to both a recombinant form of the S. aureus Fn-binding MSCRAMM rFnBD-D and to the Fn-binding synthetic peptide D2, although the titer and specificity of these antisera were good, and the IgG isolated from both immune sera recognized their respective antigens in a Western blot, the isolated IgG from the immune sera did not inhibit the binding of S. aureus to Fn.
Interestingly, antibodies purified from mice immunized with a β-D-galactosidase Fn-binding MSCRAMM fusion protein did not block the binding of S. aureus to 125I-Fn (Ciborowski et al., 1992). However, antibodies raised against the β-D-galactosidase Fn-binding MSCRAMM fusion protein that had been treated with formalin possessed moderate blocking activity. These findings suggest that modification of the antigen's protein structure was a critical factor in determining its ability to generate blocking antibodies.
It is possible that the S. aureus Fn-binding MSCRAMM interacts with Fn via an “induced fit” binding mechanism. Data supporting the concept of an “induced fit” mechanism of Fn binding by Fn-binding MSCRAMMs has recently been obtained by both immunological (Speziale et al., 1996) and physical biochemical (House-Pompeo et al., 1996) techniques.
A monoclonal antibody 3A10, isolated from a mouse injected with the S. dysgalactiae Fn-binding MSCRAMM FnBA, was recently identified and characterized. The epitope for 3A10 was localized to a newly-identified Fn binding motif (designated Au) just upstream of the repeat domain of the primary ligand binding site on FnBA. The antibody 3A10 enhanced Fn binding to Au rather than inhibiting the binding. This effect was demonstrated in two different assay systems. First, 3A10 elevated the ability of the Au-containing proteins and synthetic peptides to compete with bacterial cells for binding to Fn. Secondly, 3A10 dramatically increased the binding of biotin-labeled forms of the Au-containing proteins to the ligand immobilized on a blotting membrane. Purified 3A10 IgG did not recognize the antigen by itself, and Fn was required for the immunological interaction between the antibody and the epitope. This induction effect of Fn was shown in both Western blot and ELISA analyses in which immobilized Au-containing molecules were probed with 3A10 with varying concentrations of Fn. These data and subsequent biophysical studies indicate that the ligand binding sites of FnBA has little or no secondary structure. However, when encountered by Fn, they appear to undergo a structural rearrangement resulting in expression of a ligand-induced binding site (LIBS). The data revealed that 3A10 specifically recognized a LIBS generated by complex formation between Fn and FnBA.
1.3 Deficiencies in the Prior Art
The emerging antibiotic resistance in many bacterial species including Staphylococcus aureus and various staphylococcal and streptococcal species is being seen as a potential threat to mankind (Begley, 1994). Particularly worrisome is the multidrug resistance of Staphylococcus aureus. Today, almost half of the staphylococcal strains causing nosocomial infections are resistant to all antibiotics except vancomycin and it appears to be only a question of time before vancomycin will become ineffective (Begley, 1994). It is in this scenario that one must consider new strategies in attempts to prevent and treat infection. A detailed knowledge of the molecular pathogenesis of infections caused by staphylococci and streptococci would provide an important base in the design of these strategies. MSCRAMMs mediating microbial adhesion to the host tissues are an attractive target in the development of new antimicrobial agents since it appears to represent a critical first step in the pathogenic process of most infections.
It is clear that while a variety of approaches to the treatment of bacterial diseases have experienced some success, the growing problems of antibiotic resistance, variability of antigens between species and in the same species through mutation of antigens, and the inefficient immune systems in young children, the elderly and other immunocompromised patients, all present difficulties that need to be overcome. Thus, there exists today an immediate need for an effective treatment for streptococcal and staphylococcal pathogens that can be used for a variety of infections in both man and animals.
Burnham and coworkers reported monoclonal antibodies (mAbs) were raised to the native D1–D4 epitope protein sequence (WO 94/18327), however the biological or possible inhibitory activity of these mAbs was not demonstrated.
Interestingly, antibodies purified from mice immunized with a β-galactosidase Fn-binding MSCRAMM fusion protein did not block the binding of S. aureus to 125I-Fn (Ciborowski et al., 1992). However, antibodies raised against the β-galactosidase Fn-binding MSCRAMM fusion protein that had been treated with formalin possessed moderate blocking activity. These findings suggest that modification of the antigen's protein structure was a factor in determining its ability to generate blocking antibodies. Formalin treatment is a general method of inactivating proteins by inducing Schiff bases in lysine residues. However, the reaction is reversible and the molecule can potentially regain functionality, and therefore negate the beneficial effects of such formalin treatment.
There are several possible reasons for these results, e.g., when an active Fn-binding MSCRAMM is injected into an animal, one may assume that the bacterial protein immediately complexes with Fn. In such a complex, the surface of the MSCRAMM representing the binding site will be occupied and not available for immunological recognition. If this argument is correct, it would represent a clever previously unknown strategy used by microbes to avoid host defense mechanisms.
The fact that it has not been possible to generate blocking antibodies to the native Fn binding proteins has precluded their use in active and passive immunization methods directed toward preventing bacterial infection. Therefore, the creation of site-specifically-modified epitopes in the Fn-binding MSCRAMMs which would lead to the creation of antigens that generate high-affinity blocking antibodies would represent a significant breakthrough in the field of infectious disease, and in particular, revolutionize the manner in which staphylococcal and streptococcal diseases are treated, thereby obviating the need for antibiotic therapies to treat bacterial infection. Likewise, the development of mAb groups which recognize multiple Fn-binding epitopes from a single FnBP would facilitate methods for inhibiting Fn binding to FnBPs using monoclonal antibodies. Methods of altering DNA sequences encoding peptide epitopes would provide a stable, reproducible means for incorporating changes in the peptide sequence, and therefore represent a superior advancement over current formalin denaturation of native proteins. The development of methods of inhibiting bacterial adhesion to host cells by inhibiting ECM component interactions with bacterial cell surface proteins would provide a marked improvement in the state of the art for treatment of such infections.